Coverage matters mechanically because the consensus estimate is the bar an earnings-report is judged against, and because many institutions cannot own a stock no analyst follows. Newly public companies gain coverage in a block when the quiet-period ends; small caps that lose coverage often lose institutional ownership with it.
Read the estimate revisions rather than the ratings. Ratings are heavily skewed to buy and change slowly; numbers move first and move the stock.
Example: eight analysts model $1.42 of full-year eps, giving consensus $1.42. Two cut to $1.20 after a weak quarter, pulling consensus to $1.37, and at a constant 22x multiple that alone is a $1.10 move in fair value.
Related: quiet-period, earnings-report, guidance, eps, ipo